Member of Technical Staff, Agent Harness

About the Company

Mirendil is a tech-first company focused on solving core bottlenecks that unlock step-change acceleration across science and technology. We are building a frontier AI research company and training our own models end-to-end, with a team including researchers and engineers from Anthropic, Google DeepMind, xAI, OpenAI, Microsoft, Apple, and MIT.

Responsibilities

• Build and innovate on the agent harness, including agent loop architecture, tool integrations, prompt scaffolding, and execution environments
• Design orchestration systems for horizontal scaling of agents, covering memory, state management, and multi-agent coordination
• Build guardrails and reliability mechanisms to make long-horizon agentic tasks robust against failures and unexpected model behavior
• Own the extension layer between models and external tools, APIs, and environments
• Develop evaluation and observability tooling to measure agent behavior and catch regressions

Requirements

• Deep understanding of model capabilities
• Ability to build efficient architectures for models and researchers
• Passion for building tools and infrastructure that enable agents to work at scale